Microsoft CEO calls for an 'emergency stop' button in AI systems

Microsoft CEO Satya Nadella has emphasized the need for strict oversight in the development of artificial intelligence. He proposes creating mechanisms that would allow AI models to be shut down at any time, comparing the concept to an emergency brake in a car. According to Nadella, people should not blindly trust AI decisions, so it is necessary to ensure system transparency, auditability, and the implementation of unified safety standards. The Microsoft executive argues that as AI systems become more powerful, it becomes increasingly critical to prevent risks, conduct independent audits, and ensure timely reporting of issues. This approach aims to minimize the risk of dangerous or erroneous actions by maintaining human control over the technology.